Fender custom shop 68 Twin reissue Nasty HF Oscillation

[nsubulysses]

OMG Illinois Capacitor is located only a few miles away from me I always thought it was an asian brand that was just called Illinois Capacitor but actually had nothing to do with here.

what a drag

Oh well at least I can still enjoy searching ebay for vintage old stancor transformers which actually seem awesome and were made here as well. Bring em back home

I have seen Illinois caps fail in fender hot rod series that are under 10 years old. Funny in contrast to also occasionally see those old 1960s caps that come in valco, silvertone, sears brand, Danelectro, Supro etc that say "guaranteed for 1 year" and they still work 50-60 years later. I forgot what the brand is called. I'm sure you have seen them. They are red

[nsubulysses]

Oh jesus if you try to look up the caps you find this -- https://www.premierguitar.com/articl...to-silvertones

Never trust this amp man he talks about replacing the 60 year old "guaranteed for one year" capacitors then shows a picture of good modern caps, and it's a picturs of IC capacitors.

I think the brand of the "old reliable" ones are Planet ?

sorry for

[Helmholtz]

Like I wrote earlier, I am quite certain the problem is situated around the grid of V4b. The previous amp's problem was corrected here while and I can remove the following tubes and still see oscillation on V4b's anode. Grounding V4a's grid and separating the cathodes here have not shown any improvement.

What changes if you lift one end of C13? Or lift one end of R30 (1M)?

[Helmholtz]

Scope setting was at ,1ms

This makes the HF oscillation frequency something like 60kHz (I count ca. 6 positive HF peaks per division, one full division corresponds to f=1/0.1ms=10kHz). Referring to pictures 2,3,4. Pic1 must be a different setting.

[BiBi]

Ok, I see where you are going with this now...

First removed the connection to C13, still looks the same.
then I removed R30, still the same.
Then, I bridged R32 and C17 (the 3M3 and 10pF combo) and it is worse! Goes directly into hysterics even earlier.

Thanks for sticking with me on this one.

[BiBi]

Now I am finding some interesting things...

So, thanks to the suggestion from Helmholtz, I lifted the connections to the reverb driver (C13) and the reverb return (R30). That isn't part of the problem.
Next, I unlinked the two anodes V1b and V2b and ran a 22n cap directly to the grid V4b (with 1M to ground). Complete chaos, even without the volume up.
Then I added a 10k to the grid, and with the volume down, I am getting a healthy 3V signal on the output. Turning up the volume brings the signal down briefly,
then bursts into oscillation again at around 3 on the volume.

Scoping the output of R11 without a cap connected is clean.

[Helmholtz]

Next, I unlinked the two anodes V1b and V2b and ran a 22n cap directly to the grid V4b (with 1M to ground). Complete chaos, even without the volume up.

What happens if you just unlink V1b and V2b plates, everything else original?
What happens if you disconnect C17, everything else original.
Can you see oscillation at the plate of V2b (pin 6).

Next, I unlinked the two anodes V1b and V2b and ran a 22n cap directly to the grid V4b (with 1M to ground). Complete chaos, even without the volume up.

No good idea to bridge or skip R32//C17 as together with R31 they constitute a (compensated) voltage divider and provide "isolation" between reverb circuit input and output.

Please always do just one modification at a time and restore to original if there's no improvement.

You might try grid stoppers for V1b and V2b.

Then I added a 10k to the grid, and with the volume down, I am getting a healthy 3V signal on the output. Turning up the volume brings the signal down briefly,then bursts into oscillation again at around 3 on the volume.

Not clear: Which grid? 3V signal or oscillation?"

[BiBi]

What happens if you just unlink V1b and V2b plates, everything else original?
What happens if you disconnect C17, everything else original.
Can you see oscillation at the plate of V2b (pin 6).

disconnecting the anodes has no effect as can be seen when I clip in a capacitor directly from V1b to V4b or remove one of V1 or V2.
same with C17, if I bypass it and there is still oscillation then it shouldn't be part of the equation anymore, or?
Oscillation occurs at V2b when everything is connected, yes. When the coupling cap is disconnected the oscillation is gone.

No good idea to bridge or skip R32//C17 as together with R31 they constitute a (compensated) voltage divider and provide isolation between reverb circuit input and output.

Yes but I wanted to see if this dry line had some effect while the reverb return line was disconnected.

Please always do just one modification at a time and restore to original if there's no improvement.

yes and no. There is reasoning to this I think, but if we always go back after taking one step, I am not sure where it is leading us.

You might try grid stoppers for V1b and V2b.

these were provided for by the factory.
Honestly, I believe that the problem lies with V4 due to a previous experience with these amps

Not clear: Which grid? 3V signal or oscillation?

I added 10k to the grid of V4b. I clipped in a 22n cap with a 1M resistor to ground to go directly from the anode of V1b to the grid of V4b. With my generator plugged into the "normal" input,
there was a 3VAC clean sine wave on the output of the amplifier. Turning the volume up brought the sine wave down briefly, then a burst of oscillation at a volume level of 3.

[Helmholtz]

Oscillation occurs at V2b when everything is connected, yes. When the coupling cap is disconnected the oscillation is gone.

This doesn't point to V4b as being the (sole) source of oscillation, especially if the amplitude of oscillation is greater before R32 than after. Oscillation often/mostly involves more than one stage/tube. A grid alone cannot produce oscillation. Sustained oscillation requires gain and some (positive) feedback loop.
(A high ESR filter cap or a poor common ground can provide such feedback coupling between stages as can lead dress or component placement.)

[Helmholtz]

disconnecting the anodes has no effect as can be seen when I clip in a capacitor directly from V1b to V4b or remove one of V1 or V2.
same with C17, if I bypass it and there is still oscillation then it shouldn't be part of the equation anymore, or?

No, not the same at all.

[BiBi]

I wasn't suggesting that V4b is the sole source of oscillation. I was trying to show the other components leading to V4b are not the source of the problem.
It points to a lead dress or grounding issue as you say. Since there were similar problems on another amp I worked on recently, then I would most likely
rule out components being an element of the problem.

[TimmyP1955]

its a feature, not a bug, as all twins have that oscillation.
As far its the resistor or capacitor of the feedback loop.

Not in my experience. Every Twin we have (both vintage and RI) can be run at any volume level without oscillation.

[Chuck H]

Not in my experience. Every Twin we have (both vintage and RI) can be run at any volume level without oscillation.

Agree. And thank you for calling out the elephant in the room. Plenty of Fender amps suffer borderline instability, but that was never the intention. I think it's a consequence of over engineering and low cost labor that buried the Fender designs in the SF era. When you have two channels, spring reverb and vibrato in a 100+W combo with two speakers you really should pay attention to what has gone before and worked. Instead CBS went with cheaper, less educated labor and rougher lead dress with bleeder cap corrections for oscillation problems. The end result was a less stable amp with lower fidelity. That's what you get when you don't educate your staff about what they're actually trying to accomplish. Sure, they're paid less. But they don't know what they're doing!?! Now you don't have a product to sell because it doesn't #*%cking work!!! That's what happened to Fender in the SF era (my unabridged opinion and perspective).

[John G]

Hi,
My export schematic has grid stoppers for Africa, Fender must of bought all of the worlds 10k resistors......EU requirement ??
The input to the reverb driver has a 100k resistor to the grid(R91)of V3 ???
Also there is a ferrite Ring on the reverb tank cable.
Al this may or may not be of help ?
John

[Helmholtz]

Hi,
My export schematic has grid stoppers for Africa, Fender must of bought all of the worlds 10k resistors......EU requirement ??
The input to the reverb driver has a 100k resistor to the grid(R91)of V3 ???
Also there is a ferrite Ring on the reverb tank cable.
Al this may or may not be of help ?
John

Would you mind to post your schematic? Up to now we only have a modified '65 TR RI schem.